EV Charging Infrastructure: Designing the Whole Site

Public charging, workplace charging, residential charging and fleet-depot charging solve different problems. A public site may prioritize discoverability, payment, queuing and broad vehicle compatibility. A workplace may prioritize long dwell times and equitable sharing. A fleet depot may care almost entirely about departure readiness, route reliability and operating cost. The same charger can be suitable in one context and wasteful in another.

Start with the vehicle-day. For every meaningful vehicle group, record battery capacity, energy consumed per route, arrival time, arrival state of charge, minimum departure state of charge, dwell window, vehicle inlet and maximum AC or DC acceptance. Add seasonal variation, exceptional routes and future fleet growth. This turns an equipment conversation into an energy-and-time problem.

Power, measured in kilowatts, is the instantaneous rate of energy transfer. Energy, measured in kilowatt-hours, accumulates over time. A charger delivering 60 kW continuously for one hour would transfer 60 kWh at its output. Real sessions include conversion losses, auxiliary consumption, vehicle-imposed limits, interruptions and charge taper, so nameplate power multiplied by elapsed time is only an upper-bound estimate.

The vehicle controls how much power it accepts. With AC charging, the vehicle’s onboard charger converts AC to DC and may be the limiting component. With DC charging, conversion occurs in the external charger, but the vehicle still negotiates voltage and current and reduces power as battery temperature or state of charge requires. The statement “a 60 kW charger charges every vehicle at 60 kW” is therefore false.

A useful first calculation is required average power: energy needed divided by the charging window. A vehicle needing 90 kWh over a six-hour dwell requires an average of 15 kW at the vehicle, before allowances for losses and operational margin. This does not mean a 15 kW charger is automatically correct. It means the service requirement has been quantified and can now be tested against schedules, peak periods and exceptions.

The energy path may include the distribution utility service, service transformer, main switchgear, revenue and submetering, branch protection, distribution panels, conductors, isolation or emergency devices, charger power electronics, connector and vehicle. Communications may link meters, chargers and a charging-management platform through local networking and the internet. Civil works keep these assets physically supported, accessible and protected.

Interface ownership matters. Who confirms available utility capacity? Who performs the protection study? Who supplies the cellular service? Who owns charger credentials? Who tests backend compatibility? Who restores service after a communications failure? A responsibility matrix should assign design, supply, configuration, testing, acceptance and continuing support for every boundary.

An existing 750 kVA transformer does not offer 750 kW for new chargers. It already serves building loads, has a real loading profile, operates at a power factor, experiences ambient and thermal conditions, and connects to equipment with defined ratings and fault duties. Capacity assessment should use measured interval demand, known coincident loads, transformer and switchgear ratings, voltage behavior, protection coordination, future expansion and utility requirements.

Ten 60 kW chargers create 600 kW of connected charging capacity. If the site has 350 kW of verified charging headroom, the unconstrained design is not acceptable merely because average utilization may be low. The project must either increase capacity, reduce equipment, or implement a controlled demand limit that still meets vehicle schedules.

FIGURE 6

Connected load and managed demand are different quantities

Ten 60 kW chargers installed

600 kW 10 × 60 kW connected

350 kW illustrative managed ceiling

240 kW illustrative scheduled period

0100200300400500600kW

Connected capacity stays installed · The controller holds the draw below the ceiling

What each charging tier does to the supply, the transformer and the civil works

Charger power steps up in tiers, and each tier changes the site in a predictable way. Four 60 kW chargers and one 240 kW two-outlet cabinet draw roughly the same total power; the difference is four board circuits against one dedicated feeder, and in both cases transformer headroom must be checked. The tiers below use Bescore’s published input currents where they exist and standard arithmetic where they do not; the arithmetic is illustrative and the utility’s own figures govern.

AC, 7 to 22 kW. A 7 kW single-phase charger draws about 30 A at 230 V; 11 kW and 22 kW three-phase chargers draw about 16 A and 32 A per phase at 400 V. These fit ordinary distribution boards with a dedicated circuit each. The limit is usually the vehicle: many accept 7 or 11 kW on AC regardless of the charger. This is the tier for homes, offices, hotels and any place a vehicle stands for hours.

DC, 30 to 60 kW. A floor or wall cabinet on a three-phase feed. Bescore publishes 47 A per phase for its 30 kW unit; a 60 kW unit draws about 90 to 95 A per phase, working from 400 V and the published 96 percent efficiency. Each needs its own circuit, protection that can detect DC leakage, a grounding (earthing) connection sized for the fault level, and a short trench. The board may need a new section but the transformer usually does not, if the site’s measured headroom allows it.

DC, 120 to 360 kW, one cabinet with two outlets. Bescore publishes 186, 279, 372 and 558 A per phase for its 120, 180, 240 and 360 kW units. Those are feeders, not circuits: dedicated switchgear, larger cable, and a transformer review, because a 360 kW charger draws about 380 kVA at its input and consumes most of a 500 kVA transformer on its own. The two outlets share the cabinet’s power. With one vehicle connected it can take up to the outlet’s current limit, published as 250 A with a short 400 A boost; with two connected the cabinet divides its power between them dynamically, so a second vehicle can cut the first vehicle’s rate, by up to half when both are asking for full power. Sharing is a feature when it is understood and a complaint when it is not.

DC, 360 kW and above, split into a power cabinet and dispensers. From this tier the architecture usually separates the rectifiers, in a power cabinet placed where there is room and ventilation, from the dispensers that stand at the parking bays and carry only the cable, the screen and the payment terminal. Bescore publishes power cabinets of 470 to 782 kg and dispensers of 100 kg for its split family, with each dispenser able to deliver up to 500 kW when the cabinet allocates it. The cabinet allocates power across the dispensers dynamically, so one vehicle can take most of it or several can share. This tier needs a dedicated transformer, in many cases a primary-metered service, trenches for the DC cables from cabinet to bays, and civil works for the cabinet pad. The distribution utility decides where a dedicated transformer or a primary-metered service begins; ask early, because the answer sets the schedule.

DC, 600 to 960 kW, for heavy vehicles. At the top of the range the cable, not the cabinet, is the limit. A standard air-cooled charging cable is rated around 250 A; carrying 400 A or more continuously needs coolant circulated through the cable and connector. Bescore publishes 250 A for the air-cooled outlet and 400 A, with a 600 A maximum, for the liquid-cooled outlet on its largest units. The arithmetic that follows is the one to remember: delivered power cannot exceed the vehicle’s battery voltage times the cable’s current limit, whatever the cabinet is rated. A 400-volt-class vehicle on a 250 A cable tops out near 100 kW; an 800-volt-class vehicle on the same cable near 200 kW; the same vehicle on a 400 A liquid-cooled cable near 320 kW. A boost current rating applies for seconds, not for the session. This is why the fleet survey asks for the battery voltage class.

The site loads nobody lists. Every charger draws standby power when idle; Bescore publishes under 30 W per unit, which across a hub adds up. Rectifiers distort current; Bescore publishes total harmonic distortion of 5 percent or less between half and full load for its European-standard chargers and a power factor of 0.99 or better for its largest units, and the transformer and the utility’s limits must accommodate the site’s total. These belong in the load study alongside the charging demand itself.

A static limit caps charging at a fixed value. A dynamic limit can respond to the site’s real-time load and use capacity that would otherwise remain idle. Within that envelope, the controller may share power equally, prioritize earliest departure, protect minimum state of charge, or optimize around tariffs and demand peaks. The correct policy depends on the operating outcome.

Managed charging is not magic capacity. It depends on accurate site measurement, reliable charger communications, current vehicle or schedule information and safe fallback behavior. If the site meter becomes stale, should chargers hold the last allocation, fall to a conservative limit, or stop? If the backend is unavailable, can local control preserve essential charging? These are design requirements, not afterthoughts.

Equipment selection includes input supply, output voltage and current envelope, connector, environmental rating, cable management, accessibility, impact protection, metering, display, emergency functions, communications hardware and service access. The vehicle population determines which connector and power envelope are useful. A charger that reaches its headline rating only at voltages above the fleet’s battery voltage may deliver less than expected.

Six connector families, one fleet survey

A connector is a plug and a language. The plug must fit the vehicle’s inlet, and the charger and vehicle must speak the same protocol to negotiate voltage, current and safety. Vehicles arrive with one of several families that do not interchange, and a charger is built for one family per outlet. Buying the wrong family is not fixed by software, and adapters between DC families are not a design basis.

The DC families. CCS2, the Combined Charging System on the European Type 2 base, is carried by most vehicles built for European and many Asian markets, and it is the family most often met on passenger EVs sold new in the Philippines. CCS1 is the same system on the North American Type 1 base. CHAdeMO is the older Japanese standard, declining in new vehicles but present in used Japanese imports. GB/T is the Chinese national standard, carried by Chinese-built buses, trucks and utility vehicles and by several electric jeepney platforms. The market patterns in this paragraph are a working assumption stated to explain why the survey matters; the survey, not the pattern, decides a site.

The AC families. Type 2 (IEC 62196-2) can carry single-phase or three-phase AC and is the AC inlet on CCS2 vehicles. Type 1 is single-phase and is the AC inlet on CCS1 vehicles. On AC the vehicle’s onboard charger sets the rate, so a Type 2 socket does not by itself mean 22 kW: many vehicles accept 7 or 11 kW.

The languages. IEC 61851-1 sets the general requirements for conductive charging and IEC 61851-23 the requirements for DC charging stations. The CCS families talk over DIN 70121 and ISO 15118; ISO 15118 adds authentication built into the vehicle, so a car can be identified and billed by plugging in. The GB/T family talks over GB/T 27930, with GB/T 18487 as the Chinese counterpart to IEC 61851. A charger that names one set of these on its datasheet is telling you which vehicles it can talk to. The Philippine National Standards that the Bureau of Philippine Standards has adopted for charging equipment are confirmed for the current edition before they are cited in a specification.

The catalog reflects the fork. Bescore publishes two charger lines: a European-standard line built around CCS2, with CHAdeMO and CCS1 shown as options on its smallest unit’s page, and a GB/T line for the Chinese standard. The GB/T line publishes no datasheet and no support for OCPP, the charger-to-backend protocol described below. Metro Power lists only the CCS2 family on this site today. A fleet of Chinese-built buses would need the other line, and whether it fits a Philippine project is an engineering question to settle before any offer, not a catalog fact.

The survey. For every vehicle model that will use the site, record seven things: the DC inlet family; the AC inlet family; the battery’s nominal voltage class, roughly 400 or 800 volts, because it caps the power a given cable current can deliver; the maximum DC current the vehicle accepts; the maximum AC power and whether the onboard charger is single-phase or three-phase; the communication standard the vehicle supports; and the number of vehicles of that model now and in three years. Where a value is unknown, say so; the vehicle’s dealer or its charging port label usually answers it. Metro Power uses the survey to decide how many outlets of each family the site needs, whether a mixed site or a standardization decision is cheaper, and which chargers to propose.

OCPP, maintained by the Open Charge Alliance, defines communication between charging stations and management systems. Version support alone does not prove that every profile, smart-charging behavior, security feature or vendor combination works. Procurement should identify required messages and use cases, certification status where applicable, backend version, tested configuration and evidence from integrated acceptance.

Data ownership and vendor exit deserve explicit treatment. Who owns session and diagnostic records? Can data be exported in a usable format? Can chargers move to another backend? Who controls certificates, SIMs, keys and administrator accounts? How long are security and firmware updates provided? A low initial platform price can create a costly operational dependency.

A charger requires more than a foundation and cable. Site design should consider vehicle movement, swept paths, accessible use, cable reach, trip hazards, drainage, flood exposure, sunlight, ventilation where relevant, impact protection, lighting, security, signage, emergency access and maintenance clearance. Trenching and reinstatement can dominate disruption and cost at an existing facility.

Electrical design by qualified professionals should address conductor sizing and voltage drop, grounding and bonding, isolation, overcurrent and residual-current protection as applicable, surge protection, short-circuit duty, selectivity, emergency disconnection, labeling and inspection. Exact requirements depend on the controlling Philippine rules, utility conditions, equipment instructions and site classification.

Protective grounding and leakage protection that can see DC

A charger’s protection list is also a specification for the supply that feeds it. Two lines in it are the ones an electrical inspector checks at energization, and the ones that most often stop a site from being switched on.

Grounding. A charger is a metal-cased appliance that a person holds in the rain, so its exposed metal is bonded to the protective grounding conductor (PE) and that conductor is continuous back to the supply’s grounding point. The way the supply is grounded has a name. In a TN-S system the neutral and the PE conductor run as separate conductors from the transformer; in TN-C-S they are combined for part of the way and separated at the installation; in a TT system the installation has its own grounding electrode independent of the supply’s. The charger’s manufacturer designs for one or more of these and says which; Bescore states a TN-S supply on one of its charger pages. Philippine low-voltage installations vary, and the arrangement at the actual supply board is confirmed by the electrical practitioner who signs the plans before the charger is ordered, because a charger built for TN-S on a TT supply needs different protection. Chargers also monitor the PE conductor and refuse to charge if it is lost, which is a safety function and not a nuisance.

Residual current protection. A residual current device trips when the current going out on the live conductor does not come back on the neutral, because the difference is leaking somewhere, possibly through a person. Devices are graded by what kind of leakage they can detect. Type AC sees only sinusoidal AC leakage. Type A also sees pulsating DC. Type B sees smooth DC as well. A vehicle’s onboard converter can leak smooth DC into the AC supply, and smooth DC saturates the sensing core of a Type A device and blinds it, so that it will not trip on a subsequent AC fault either. For that reason the charging standard requires, at every connecting point, either a Type B device or a Type A device combined with a separate detector that trips at 6 mA of DC leakage. That protection lives either inside the charger or in the upstream board, and the upstream board must be designed to match. Bescore lists a Type A residual current breaker in the protection list of its European-standard chargers; whether a 6 mA DC detector accompanies it is not stated on the datasheets reviewed, so it is a question to put to the supplier. On a DC charger the vehicle side is separated from the AC supply by the charger’s converter, and the DC side is protected by insulation monitoring rather than by a residual current device; the AC input still needs the protection above.

Surge. Chargers stand outdoors, often in open car parks, in a country with frequent lightning. Surge protection at the board (a Type II device, with Type I where the site is exposed) protects the charger’s electronics and the vehicle connected to it. Bescore lists surge protection on its chargers; the coordination with the board’s own devices is part of the electrical design.

Who decides. The Philippine Electrical Code and the local building official’s electrical inspector control these requirements, and the electrical practitioner who signs the plans is responsible for meeting them. Metro Power’s part is to give that practitioner the charger’s published grounding assumption, its protection list and the questions above before the design is frozen, not at the inspection.

FIGURE 9

A usable bay protects people, equipment and operations

FootpathBayNext bay123456

1. 1Cable reachthe connector reaches the vehicle inlet without strain or a trip hazard
2. 2Bollardprotects the charger from a vehicle that overruns the bay
3. 3Drainagecarries water away from the equipment, cable and connector
4. 4Lightinglets a user see the connector, markings and route at night
5. 5Accessible routea clear, level path between the bay and the charger
6. 6Service clearanceworking space to open and maintain the unit

Planning view · Not a construction drawing · Not to scale

The Electric Vehicle Industry Development Act and current Department of Energy issuances establish the national EV framework. Project obligations can depend on whether the site is private or commercially accessible, who operates the charging service, what equipment is installed and which local and distribution-utility processes apply. Electrical, building, fire, product conformity, business and local permitting questions should be verified for the actual site and current rules.

A regulatory pathway is a dependency map. It should name the question, responsible party, authority, controlling issuance, required evidence, timing and status. It should distinguish a verified requirement from an assumption or an item awaiting authority confirmation. A generic checklist cannot establish approval, compliance or energization.

Whole-site capital cost may include chargers, transformer or service upgrade, switchgear, protection, panels, conductors, trenching, foundations, bollards, network equipment, design, permits, testing, commissioning, training, initial spares, tax, freight and contingency. Operating cost may include purchased energy, demand charges, software, connectivity, payment fees, rent or revenue share, preventive maintenance, corrective maintenance, customer support, insurance and financing.

Public charging revenue depends on sessions, energy per session, price, uptime and payment collection. Fleet value may come from avoided fuel and maintenance cost, route availability, controlled demand and operational resilience rather than retail charging revenue. The commercial model should be aligned with the site’s real purpose.

A pilot should answer uncertainties that matter to scale: vehicle compatibility, delivered energy, site demand, charge scheduling, communications reliability, user behavior, payment flow, alarm response, maintenance effort and data quality. It needs a baseline, defined test population, pass criteria, observation period, issue log and decision gate.

Commissioning should verify identity and ratings, installation, protection, grounding, electrical tests, charger functions, emergency behavior, load-management limits, communications-loss fallback, authorization, transaction records, payment where applicable, reporting, fault recovery and the agreed vehicle set. Handover should include as-builts, settings, credentials, warranties, training, maintenance plan, support escalation and spare strategy.

A controlled specification separates mandatory requirements, evaluated preferences and future options. It states the use case, service level, electrical ratings, vehicle and connector scope, environmental conditions, communications, software functions, cybersecurity expectations, data ownership, warranty, spares, training, documentation, tests and continuing support. Every claim should resolve to exact product evidence and an acceptance method.

Before a commitment, classify the facts. Verified facts have a traceable source. Assumptions are visible and sensitivity-tested. Missing data has an owner and deadline. Manufacturer confirmation identifies the exact model, firmware and configuration. Management approval covers commercial exposure and residual risk. This discipline prevents the charger quotation from becoming an accidental promise for the entire site.